Although hormonal therapy produces objective evidence of remission in approximately 80% of patients with metastatic carcinoma of the prostate, the duration of this response is quite variable. However, if it were possible to predict the duration of response to endocrine manipulation, then these patients who were unlikely to achieve long term benefit could be selected for adjuvant forms of treatment at an earlier time. In a similar way, if it were possible to predict the disease free interval following radical prostatectomy or to select those patients most likely to respond to chemotherapy, this information would have great value. This study proposes to investigate the relationship between various target organ biochemical events and the biological behavior of prostatic cancer. Specifically, we propose to determine whether the measurement of steroid receptor content (androgen, estrogen, progesterone), steroid content (testosterone, dihydrotestosterone, 3 alpha-androstanediol) or enzymatic profiles will be useful in predicting: 1) the quality and duration of response to hormonal therapy in men with metastic prostatic cancer; 2) the response to chemotherapy in men with relapsing prostatic carcinoma; and 3) the disease free interval in men following radical prostatectomy.